Parallel explicit and implicit mixed discontinuous Galerkin finite element time domain method

Abstract

The invention discloses a parallel explicit and implicit mixed discontinuous Galerkin finite element time domain method. A cause and effect area decomposition method is used for realizing the parallel explicit and implicit mixed discontinuous Galerkin finite element time domain method, and highly parallel calculation can be carried out while a time step size is amplified. By use of the method, the simulation calculation time of a finite element time domain method can be effectively saved, and the method is simple in programming and has a high practical engineering application value.

Description

technical field [0001] The invention belongs to the numerical calculation technology of the explicit and implicit time-domain finite element algorithm, belongs to the field of large-scale parallel computing, and is specifically a high-efficiency algorithm for multi-scale electromagnetic analysis. Background technique [0002] Multi-scale problems appear in many practical electronic engineering fields, such as complex system electromagnetic interference, electromagnetic compatibility and communication channel analysis. Due to the existence of fine structure, the traditional time-domain finite element method needs to use small-scale subdivision mesh to simulate the physical model. At the same time, due to the limitation of stability conditions, the time iteration step size is too small, which leads to serious time-consuming system simulation. Although the selection of the time step in the unconditionally stable difference scheme is not limited by the subdivision size, it needs...

AI Technical Summary

Problems solved by technology

[0003] Although the traditional explicit discontinuous Galerkin time-domain finite element method satisfies natural parallelism, it cannot enlarge the time step due to the limitation of stability conditions

The unconditionally stable and discontinuous Galerkin time-domain finite element method, such as the implicit difference method based on CN, can enlarge the time step without being limited by the stability condition, but it needs to invert a large sparse whole in the solution process, and the limitation Therefore, although the traditional explicit and implicit combination method can effectively solve multi-scale problems, it cannot achieve high parallelism.

Images